Codebreaker Labs secured $4.5 million in seed funding, announced in September 2024, money the Boulder startup intends to spend building what its founders describe as a critical missing ingredient in genomic medicine: hard experimental evidence about what genetic mutations actually do inside living human cells. Kickstart led the round, joined by Buff Gold Ventures, Children's Hospital Colorado, Denver Ventures, Service Provider Capital and Pelican Trust.
The company plans to release its first commercial dataset by 2027, a timeline that reflects both the ambition and the logistical complexity of its approach. Unlike the algorithmic shortcuts many genomics companies have embraced, Codebreaker is running physical experiments at scale, installing tens of thousands of genetic variants into disease-relevant cells using pooled CRISPR techniques, then measuring the downstream effects with single-cell RNA sequencing.
It's painstaking work. Of the roughly 9 billion possible single-letter mutations across the human genome, according to Codebreaker's materials, fewer than 0.05 percent have been validated in the lab. That leaves a sprawling territory of genetic uncertainty, particularly for the variants flagged as "of uncertain significance" in clinical sequencing reports, a designation that often leaves patients and physicians with more questions than answers.
An August 2024 securities filing disclosed a larger target: the company's SEC Form D shows an offering of approximately $7.4 million with roughly $5.9 million sold as of that filing date.
Building Atlases, One Variant at a Time
Codebreaker's immediate plans involve scaling its laboratory operations, hiring scientists and computational biologists, and assembling what it calls variant "Atlases." These disease-specific datasets marry experimental measurements with AI interpretation tools, offering what the company frames as a more grounded alternative to purely computational predictions.
The platform begins with computational methods that mine variant databases including the GWAS Catalog and ClinVar. Researchers then engineer those single-nucleotide changes into primary human cells tailored to the disease in question. T cells for immunology studies, for instance. The edited cells undergo single-cell RNA sequencing, generating readouts that machine learning models use to score variants whose clinical significance remains murky.
According to the company's website, a single experimental run can handle tens of thousands of variants and produce a queryable Atlas within three months. The first such Atlas, targeted for 2027, aims to measure 100,000 genomic variants tied to immunology and blood cancers. The company lists three Atlases currently under development, focused on pediatric immunology, pediatric leukemia and adult leukemia. Additional projects addressing inflammatory bowel disease, rheumatoid arthritis and systemic lupus erythematosus remain in design phases.
The Inscripta Connection
CEO and co-founder Ryan T. Gill brings a track record in the genome-editing sector. He previously co-founded Inscripta, serving as founding CEO and chief scientific officer before launching another venture, Artisan Bio. Co-founder and CTO Tanya Warnecke-Gill worked alongside him at both companies and holds more than 35 patents, per company materials. The third co-founder, Ryan Layer, is an assistant professor of computer science at the University of Colorado Boulder, where his research focuses on structural genomic variation and large-scale algorithmic approaches.
The board includes Dalton Wright, a general partner at Kickstart, and Mark Lupa, co-founder and general partner at Buff Gold Ventures. Sandy Zweifach, listed in SEC filings as a director and identified in the press release as company chairman, rounds out the leadership team.
LinkedIn data suggests the startup currently employs somewhere between two and ten people, though such self-reported figures may not reflect current headcount and the company is actively hiring. Posted positions include a director of biology with a salary range of $145,000 to $190,000, a biology scientist role at $90,000 to $115,000, and a computational biologist position offering $120,000 to $190,000. Job descriptions reference technical methods such as "pooled HDR variant installs" and "variant Perturb-seq," which layers CRISPR perturbations over single-cell sequencing.
Partnerships and the Competitive Landscape
Codebreaker has announced collaborations with Children's Hospital Colorado to develop and validate its AI variant-interpretation tools. The hospital also invested in the seed round. Parse Biosciences is providing single-cell sequencing technology to work with Codebreaker's variant libraries, while Twist Bioscience supplies the oligonucleotide pools that underpin those libraries.
The broader competitive terrain is somewhat unusual. Experimental variant-effect mapping has traditionally been the province of academic consortia and public repositories. MaveDB, a publicly accessible database, reportedly holds more than 7 million variant-effect measurements as of recent updates. The Atlas of Variant Effects Alliance coordinates academic labs working on similar mapping efforts.
On the predictive front, companies and research groups have turned to AI models trained primarily on sequence data and statistical correlations from genome-wide association studies. Some efforts claim to precompute effects for billions of possible variants using purely computational methods, though such predictions lack the cell-based experimental validation Codebreaker is pursuing.
"Genomic AI needs more than DNA sequences," Gill said in the announcement. "It needs to learn from what genetic changes actually do in human cells."
Wright, the Kickstart investor, framed the bet in market terms. "High-quality experimental data will become an increasingly valuable asset" as the genomic AI sector matures, he noted.
Whether that asset proves valuable enough to justify the slower, more capital-intensive path Codebreaker has chosen remains an open question. The company is wagering that in a field awash with predictions, verified biological truth will command a premium.
